Repeatedly Solving Similar Simulation-Optimization Problems: Insights from Data Farming

We study a setting in which a decision maker solves a series of similar simulation-optimization problems, where parameters of the problem vary over time to reflect the most up-to-date conditions. For example, in a generalized newsvendor problem, the newsvendor must choose how much of each raw material to order each day in response to the latest prices for procuring and salvaging raw materials. At the beginning of each day, a simulation-optimization solver is used to recommend a solution for that day's problem. This setting prompts several questions of potential interest: How sensitive are the solver's recommended solution and performance to changes in the problem's parameters? Does the use of a fixed set of solver hyperparameters lead to consistently good performance, or would judiciously tuning these hyperparameters lead to a practically significant improvement? We leverage data farming to provide some preliminary answers and describe additional insights uncovered during the process.